Hypothesis 1: Stereotype Taxonomy. Results from the stereotype ratings provided
by the Identity Developmental Laboratory were analyzed using inter-rater reliability methods with consistency agreement. This was done to determine the underlying
taxonomy of student-athlete stereotype traits, with the consistency measure of inter-rater agreement tapping into the extent to which raters were able to access a shared working understanding of stereotypes. Due to low ratings of understandability by the
undergraduate research assistants, 144 words were removed before conducting the analysis. Based on the overall distribution of the understandability ratings, any word that resulted in than 4.3 on a scale of 1 – 5, with 1 being “Never seen this word before in my life” and 5 being “Can use it successfully in a sentence,” was removed. Results from the inter- rater reliability analyses concluded highly consistent ratings between the eight members of the Identity Development Lab and the three additional research assistants. Intraclass correlation coefficients ranged from 0.740 (stereotypicality of White individuals ratings of traits) to 0.974 (likablity ratings of traits). See Table 1 for a summary of the results.
Table 1. Intraclass Correlations for Consistency of Trait Adjective Ratings RA ratings Rating Scale Intraclass Correlation
Student-athletes 0.910
Non-athlete students 0.749
White 0.740
Black 0.881
Likability 0.974
Comparing these to the intraclass correlations from Anderson (2015) indicates consistency across different raters over different years using the same method (student- athletes = 0.824; non-athlete students = 0.735; likability = 0.968).
Cutoffs were placed to determine the final list of person-adjectives. The likability cutoffs, based on Anderson’s (1968) 1 – 7 scale, were as follows: positive traits were rated between 5.0 and 7.0; neutral traits were rated between 3.0 and 4.9; and negative traits were rated between 1.0 and 2.9. Squared deviations between student-athlete
stereotypicality ratings, non-athlete student stereotypicality ratings, White stereotypicality ratings, and Black stereotypicality ratings were used to determine the 18 traits rated as most stereotypical of each category. 39 of the current traits were overlapping from Anderson (2015).
Hypothesis 2: Non-Random Assignment to Categories. One-sample chi-square
tests were run to determine if participants were assigning traits in a non-random manner. Significant results at this level indicate that the sorting was not random and that
participants used categories differently with different traits. The chi-square tests used the Bonferroni adjusted significance values (p < 0.0007). See Appendix I for significance tests.
Hypothesis 3a: Racial Identification and Athlete Status – Forced-Choice. Results
were organized based on several categories: traits associated with race and athlete status (White student-athlete, Black student-athlete, White non-athlete student, and Black non- athlete student); traits associated with either race or athlete status (White, Black, student- athlete, non-athlete); and traits not associated with race and athlete status (not Black student-athlete, not White student-athlete, not Black non-athlete student, not White non- athlete student). Organization of categories will be decided based on the proportion of participants who placed a trait into that category. For example, if a trait was placed into Black student-athlete 77% of the time, and the next closest group was Black non-athlete student at 11% of the time, this trait was assigned to the Black student-athlete category.
Traits that have similar percentages (less than or equal to a 10% difference) in two
categories was placed into the more general category. For example, if a trait was placed in the White student-athlete category by 56% of the participants or in the Black student- athlete category by 47% of the participants, this trait was assigned to the general athletic category. Additionally, traits that were found to be not often associated with one or more of the categories were placed in a group containing the word “not” before the category it does not describe (e.g., not White student-athlete). If one group had less than or equal to 10% of participants, and the next lowest group had less than 20% or participants, the trait was placed into this particular category. For example, if a trait was placed in the Black Student-Athlete category by 31% of participants, placed in the White Student-Athlete category by 29% of participants, placed in the Black Non-Athlete Student category 32% of the time, and placed in the White Non-Athlete Student category 8% of the time, the trait was assigned to the not White Student-Athlete category. The complete set of results are presented in Appendix J.
Black Student-Athlete. Eight of the adjectives assessed (11.1%) were associated
with Black student-athletes. These adjectives were aggressive, amusing, clownish, comical,
dominating, entertaining, hot-headed, and proud. Three traits were positive, two were
Table 2. Black Student-Athletes Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Aggressive 0.57 0.19 0.22 0.03 202.077 Amusing 0.48 0.18 0.29 0.06 124.834 Clownish 0.43 0.18 0.31 0.09 87.911 Comical 0.47 0.14 0.34 0.06 136.182 Dominating 0.69 0.20 0.08 0.03 355.923 Entertaining 0.67 0.11 0.19 0.04 314.077 Hot-Headed 0.53 0.28 0.17 0.02 108.957 Proud 0.51 0.36 0.10 0.03 198.582
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
White Student-Athlete. Seven of the adjectives assessed (9.7%) were associated
with White student-athletes. These adjectives were attentive, conceited, disciplined, humble,
reasonable, respectful, and vain. Four traits were positive, one was neutral, and two were
Table 3. White Student-Athlete Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Attentive 0.23 0.35 0.22 0.20 17.462 Conceited 0.34 0.46 0.15 0.06 126.582 Disciplined 0.30 0.50 0.10 0.11 136.009 Humble 0.19 0.37 0.25 0.19 28.145 Reasonable 0.16 0.39 0.23 0.23 35.948 Respectful 0.18 0.40 0.19 0.24 42.126 Vain 0.27 0.39 0.23 0.11 49.658
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Black Non-Athlete Student. Three of the adjectives assessed (4.2%) were
associated with Black non-athlete students. These adjectives were rebellious, unhappy, and
unpleasing. One trait was positive and two were negative. Results are presented in Table 4.
Table 4. Black Non-Athlete Student Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Rebellious 0.30 0.12 0.52 0.06 171.111
Unhappy 0.06 0.11 0.61 0.23 243.037
Unpleasing 0.16 0.27 0.45 0.13 79.788
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
White Non-Athlete Student. Sixteen of the adjectives assessed (22.2%) were
associated with White non-athlete students. These adjectives were antisocial, boring,
clumsy, gentle, innocent, literary, mathematical, ordinary, philosophical, quiet, scientific, sensitive, shy, soft-spoken, sophisticated, and timid. Six traits were positive, eight were
neutral, and two were negative. Results are presented in Table 5.
Table 5. White Non-Athlete Student Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced- Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Antisocial 0.06 0.10 0.36 0.48 161.511 Boring 0.04 0.15 0.19 0.63 266.668 Clumsy 0.06 0.15 0.28 0.51 149.055 Gentle 0.08 0.16 0.25 0.52 147.209 Innocent 0.06 0.24 0.19 0.51 140.194 Literary 0.05 0.19 0.22 0.54 164.834 Mathematical 0.05 0.13 0.11 0.72 392.378 Ordinary 0.08 0.15 0.20 0.56 179.874 Philosophical 0.09 0.13 0.26 0.53 154.938 Quiet 0.07 0.15 0.31 0.48 127.468 Scientific 0.03 0.09 0.16 0.72 393.728 Sensitive 0.08 0.16 0.33 0.44 101.794 Shy 0.04 0.12 0.32 0.52 178.865 Soft-Spoken 0.08 0.17 0.29 0.46 100.908 Sophisticated 0.11 0.30 0.15 0.43 87.394 Timid 0.06 0.15 0.32 0.47 134.926
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Black. Two of the adjectives assessed (2.8%) were Black-identified traits. These
adjectives were hostile and impulsive. One trait was neutral and one was negative. Results are presented in Table 6.
Table 6. General Black Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Hostile 0.35 0.18 0.42 0.06 100.022
Impulsive 0.43 0.19 0.33 0.05 109.548
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
White. Four of the adjectives assessed (5.6%) were White-identified traits. These
adjectives were calm, clean, middle-class, and well-mannered. Three traits were positive and one was neutral. Results are presented in Table 7.
Table 7. General White Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Calm 0.12 0.28 0.23 0.38 44.883
Clean 0.10 0.38 0.12 0.40 101.695
Middle-Class 0.10 0.31 0.19 0.41 72.625
Well-Mannered 0.14 0.35 0.17 0.34 47.320
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Athletic. Six of the adjectives assessed (8.3%) were associated with athletes in
general. These adjectives were egotistical, experienced, persistent, popular, self-assured, and
skillful. Three traits were positive, two were neutral, and one was negative. Results are
presented in Table 8.
Table 8. General Student-Athlete Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Egotistical 0.42 0.44 0.09 0.06 162.274 Experienced 0.38 0.45 0.10 0.07 142.926 Persistent 0.43 0.38 0.14 0.06 124.342 Popular 0.49 0.48 0.02 0.01 290.102 Self-Assured 0.36 0.45 0.12 0.07 133.843 Skillful 0.67 0.27 0.04 0.04 335.345
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Non-Athletic. Four of the adjectives assessed (5.6%) were associated with non-
athletes in general. These adjectives were artistic, reserved, soft-hearted, and thoughtful. Three traits were positive and one was negative. Results are presented in Table 9.
Table 9. General Non-Athlete Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Artistic 0.07 0.06 0.44 0.44 181.843
Reserved 0.11 0.19 0.37 0.34 42.126
Soft-Hearted 0.15 0.22 0.28 0.35 28.982
Thoughtful 0.14 0.22 0.29 0.36 36.366
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Traits Not Assigned to a Category. One of the adjectives assessed (1.4%) did not
have a significant one-sample chi-square using the Bonferroni adjustment. Therefore, it could not be differentiated across categories. This adjective was modern. Results are presented in Table 10.
Table 10. Non-Categorized Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Forced-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Modern 0.31 0.26 0.23 0.20 8.428
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
White Student-Athlete and Black Non-Athlete Student. One of the adjectives
assessed (1.4%) was placed within the White student-athlete and Black non-athlete student categories equally: fault-finding.
Not Black Student-Athlete. Eight of the adjectives assessed (11.1%) were
consistently not associated with Black student-athletes. These adjectives were cautious,
dull, intellectual, jealous, realistic, sensible, smart, and warm.
Not White Student-Athlete. One of the adjectives assessed (1.4%) was consistently
not associated with White student-athletes: theatrical.
Not Black Non-Athlete Student. No adjectives were associated with this category. Not White Non-Athlete Student. Eleven of the adjectives assessed (15.3%) were
consistently not associated with White non-athlete students. These adjectives were
boisterous, crude, excitable, forceful, immature, impolite, lively, moody, obnoxious, sociable,
and unintelligent.
Hypothesis 3b: Racial Identification and Athlete Status – Free-Choice. Similarly to
part a of Hypothesis 3, results were organized based on several categories: traits associated with race and athlete status (White student-athlete, Black student-athlete, White non- athlete student, and Black non-athlete student); traits associated with either race or athlete status (White, Black, student-athlete, non-athlete); and traits not associated with race and athlete status (not Black student-athlete, not White student-athlete, not Black non-athlete student, not White non-athlete student). All proportions for the free-choice sort can be found in Appendix K.
Black Student-Athlete. Six of the adjectives assessed (8.3%) were associated with
Black student-athletes. These adjectives were amusing, dominating, entertaining, forceful,
hot-headed, and skillful. Three of the traits were positive and three were negative. These
Table 11. Black Student-Athletes Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Free-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Amusing 0.45 0.15 0.30 0.10 98.15 Dominating 0.62 0.27 0.05 0.06 275.55 Entertaining 0.68 0.11 0.18 0.03 336.60 Forceful 0.46 0.30 0.17 0.07 109.15 Hot-Headed 0.58 0.23 0.15 0.03 217.73 Skillful 0.51 0.27 0.12 0.11 138.59
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
White Student-Athlete. Four of the adjectives assessed (5.6%) were associated with
White student-athletes. These adjectives were conceited, disciplined, self-assured, and vain. Two of the traits were positive and two were negative. Results are presented in Table 12. Table 12. White Student-Athlete Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Free-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Conceited 0.23 0.55 0.07 0.15 169.63
Disciplined 0.28 0.51 0.07 0.14 142.75
Self-Assured 0.26 0.50 0.08 0.17 123.68
Vain 0.27 0.40 0.19 0.13 53.13
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Black Non-Athlete Student. Two of the adjectives assessed (2.8%) were associated
with Black non-athlete students. These adjectives were rebellious and unpleasing. One of the traits was positive and one was negative. Results are presented in Table 13.
Table 13. Black Non-Athlete Student Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Free-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Rebellious 0.25 0.15 0.49 0.12 111.05
Unpleasing 0.12 0.21 0.44 0.23 70.56
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
White Non-Athlete Student. Thirty of the adjectives assessed (41.6%) were
associated with White non-athlete students. These adjectives were antisocial, artistic,
attentive, boring, cautious, clean, clumsy, dull, gentle, innocent, intellectual, literary, mathematical, middle-class, ordinary, philosophical, quiet, realistic, reasonable, reserved, respectful, scientific, sensitive, smart, soft-hearted, soft-spoken, sophisticated, thoughtful, warm, and well-mannered. Seventeen traits were positive, nine were neutral, and four were
Table 14. White Non-Athlete Student Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Free-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Antisocial 0.04 0.06 0.25 0.66 323.63 Clean 0.07 0.35 0.11 0.47 146.62 Intellectual 0.06 0.15 0.17 0.63 260.64 Middle-Class 0.08 0.28 0.15 0.49 127.57 Ordinary 0.06 0.10 0.13 0.72 391.20 Realistic 0.12 0.27 0.16 0.46 87.96 Reserved 0.07 0.15 0.26 0.53 159.10 Smart 0.07 0.17 0.17 0.59 210.50 Soft-Hearted 0.10 0.14 0.27 0.51 132.71 Thoughtful 0.10 0.15 0.24 0.52 138.20 Warm 0.07 0.16 0.26 0.51 136.53 Well-Mannered 0.14 0.25 0.15 0.46 82.15
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Black. Four of the adjectives assessed (5.6%) were Black-identified traits. These
adjectives were clownish, comical, hostile, and impulsive. One of the traits was positive, one was neutral, and two were negative. Results are presented in Table 15.
Table 15. General Black Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Free-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Clownish 0.35 0.18 0.36 0.12 58.52
Comical 0.39 0.14 0.41 0.07 115.36
Hostile 0.28 0.23 0.38 0.11 49.51
Impulsive 0.29 0.27 0.35 0.09 48.63
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
White. One of the adjectives assessed (1.4%) was a White-identified trait: humble.
This trait was assessed as positive. Results are presented in Table 18.
Table 18. White Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Free-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Humble 0.18 0.33 0.21 0.28 16.65
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Athletic. Seven of the adjectives assessed (9.7%) were associated with athletes in
general. These adjectives were boisterous, egotistical, experienced, persistent, popular,
proud, and sociable. Two of the traits were positive, four were neutral, and one was
Table 16. General Student-Athlete Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Free-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Boisterous 0.44 0.31 0.19 0.06 104.01 Egotistical 0.37 0.38 0.11 0.15 80.70 Experienced 0.35 0.42 0.11 0.12 97.22 Persistent 0.41 0.34 0.13 0.13 82.47 Popular 0.46 0.47 0.03 0.05 240.03 Proud 0.42 0.44 0.08 0.07 161.76 Sociable 0.38 0.34 0.12 0.11 65.95
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Non-Athletic. Four of the adjectives assessed (5.6%) were associated with non-
athletes in general. These adjectives were moody, shy, timid, and unhappy. Two of the traits were neutral and two were negative. Results are presented in Table 17.
Table 17. General Non-Athlete Stereotype Trait-Adjectives Based on Proportions of Participants Assigning Trait Stereotypes to Group Categories – Free-Choice
Student-Athletes Non-Athlete Students
Trait Black White Black White
χ
2Moody 0.20 0.17 0.33 0.30 22.31
Shy 0.02 0.08 0.28 0.62 285.82
Timid 0.03 0.07 0.30 0.60 262.02
Unhappy 0.02 0.09 0.42 0.47 198.31
Note. Sample Size = 324. Chi-square ( 2) values above 11.5 are statistically significant (p <.0007).
Not Black Student-Athlete. Four of the adjectives assessed (5.6%) were
consistently not associated with Black student-athletes. These adjectives were calm, fault-
finding, jealous, and sensible.
Not White Student-Athlete. One of the adjectives assessed (1.4%) was consistently
not associated with White student-athletes: theatrical.
Not Black Non-Athlete Student. One of the adjectives assessed (1.4%) was
associated with this category: modern.
Not White Non-Athlete Student. Eight of the adjectives assessed (11.1%) were
consistently not associated with White non-athlete students. These adjectives were
aggressive, crude, excitable, impolite, impulsive, lively, obnoxious, and unintelligent. Comparison of Methods: Multidimensional Scaling
The extent to which there may be underling structural differences in how
participants respond to both the free- and fixed- choice stereotype tasks is an additional component of Hypothesis 3. To evaluate the overall structure of stereotype assignment, non-metric Multidimensional Scaling (MDS; Kruskal & Wish, 1978) was used to develop a spatial representation of the inter-relations between trait adjectives. In general, the MDS technique can be used to represent the inter-relations among a set of characteristics (based on some measure of similarity) in terms of their relative proximity in a structure defined as a set of dimensions. When analyzing the card-sort data, it is possible to measure the degree of similarity between trait adjectives using a co-occurrence matrix, with numeric values representing the proportion of times traits were assigned to the same category across participants(Coxon, 1999), with a range from 0 (traits never assigned to the same category) to 1.0 (traits always assigned to the same category.
The analysis of co-occurrence for the trait assignment data produces two
symmetrical 72x72 matrices, one representing the co-occurrence of traits under the forced sort condition, and the second representing the free sort condition. MDS provides a
potential framework for understanding and interpreting the overall pattern of stereotypes, by analyzing variability in trait assignment to stereotype categories in terms of underling dimensions. The fit of the data to the MDS solution is assessed using two indices, a Stress value, which can range from 0 to 1 (with lower values indicating a better fit) and an estimate of Variance Accounted For (VAF) by the model, which can also range from 0 to 1 (but with higher values indicating a better fit).
The MDS for the in-lab and online assessments can be found in Appendix L. The graphical representations of the data clearly show a difference between methods, and that the in-lab forced-choice data has more defined dimensions. Based on the words associated with each part of the graphic, it was determined that Dimension 1 is classified as “Athlete Status” and Dimension 2 is classified as “Valence” (positive, negative, neutral) word types. The upper right quadrant represents “Negative Non-Athletes,” (e.g., Antisocial) the lower right quadrant represents “Positive Non-Athletes,” (e.g., Sensible) the upper left quadrant represents “Negative Athletes,” (e.g., Crude) and the lower left quadrant represents “Positive Athletes,” (e.g., Persistent).
The same is not true of the free-choice data: the data is much messier with more traits clustered around one side of the figure. Dimension 1 is classified as “Athletic Status” and Dimension 2 is classified as “Valence.” The upper right quadrant represents “Negative Non-Athletes,” (e.g., Moody) the lower right quadrant represents “Positive Non-Athletes,” (e.g., Clean) the upper left quadrant represents “Negative Athletes,” (e.g., Vain) and the
lower left quadrant represents “Positive Athletes,” (e.g., Popular). It appears that, when asked to sort traits without the limitation of equal number per category, participants tend to place them disproportionately into the Non-Athlete categories. This could be an artifact of the struggles inherent in the original choices of trait adjectives. When participants are asked to sort traits with the limitation of an equal number of traits per category, it may make the data somewhat cleaner on a structural level. However, this ‘cleanness’ may be a product of the structured nature of the task. More structure may result in more spillover, because participants could not put all the “non-athlete” traits into a non-athlete category, making the data look cleaner while actually hiding some of the nuances within the data.
Hypothesis 4: Sex Differences in Stereotype Assignment – Participant Effects. A
series of 2X4 (sex of participant X White student-athlete/Black student-athlete/White non- athlete student/Black non-athlete student) chi-square tests of association were run to determine if the sex of the participant impacted their placement of cards. Significant values at the Bonferroni adjusted level (p < 0.0007) indicate that the sex of the participant